An AI medicine delivery robot

Abstract

The utility model provides a kind of AI medicine delivery robot, including robot body and multi-wheel adaptive damping chassis;Robot body is built-in central processing unit, can construct navigation path in real time by environmental perception module, and control chassis movement;Scan medicine box identification code and patient information through visual recognition unit, verify drug matching;Trigger early warning or path re-planning through communication module. Multi-wheel adaptive damping chassis includes at least two groups of rocker structure and is arranged at the bottom of robot body through shock absorber;Robot body is loaded with medicine cabinet and medicine taking arm: medicine cabinet includes multiple liftable independent medicine shelves and magnetic medicine box, each medicine box is provided with unique identification code and magnetic adsorption component;The end of medicine taking arm is provided with visual recognition unit and magnetic suction actuator, for accurate grabbing medicine box;The utility model effectively solves the problems of drug mismatch, low efficiency, strong artificial dependence and unstable drug transportation of existing medicine delivery robot in medicine delivery scene.

Description

Technical Field

[0001] This utility model relates to the field of medical auxiliary devices and intelligent robots, and in particular to an AI-powered drug delivery robot. Background Technology

[0002] In modern medical and elderly care settings, medication delivery often faces manpower shortages. Nurses frequently travel between pharmacies and wards, consuming significant time and energy, greatly increasing their workload. While traditional medication delivery robots can alleviate manpower pressure to some extent, they still have obvious shortcomings. In the medication retrieval and placement process, existing robots lack effective monitoring and error correction mechanisms. Nurses are prone to errors when storing medications due to repetitive work, leading to mismatches between prescribed medications and patient diagnoses. Existing robots lack the ability to determine correct medication placement, which not only affects patient treatment outcomes but may also cause serious medical accidents, resulting in secondary harm to patients and increasing the risk of medical disputes. Furthermore, in complex environments such as hospitals and nursing homes where frequent medical care is required, medication delivery robots must navigate various obstacles, such as thresholds and anti-slip surfaces. The chassis design of existing robots has limitations in terms of maneuverability; four-wheel drive can easily cause bumps or jams during medication transport, affecting the stability of the delivery. This is especially true for certain dosage forms such as injections and capsules, where bumps may alter their physical properties and affect efficacy. Utility Model Content

[0003] The purpose of this invention is to solve the problems of drug mismatch, low delivery efficiency, high dependence on manual labor, and unstable drug transportation by existing drug delivery robots in the current medical drug delivery scenario, and to provide a more efficient, safe, and intelligent drug delivery solution.

[0004] To achieve the above objectives, this utility model proposes an AI-powered medicine delivery robot, comprising a robot body and a multi-wheeled adaptive shock-absorbing chassis:

[0005] The robot body has a built-in central processing unit and integrates a display screen, an environmental perception module, a voice module, a positioning module, a wireless communication module, a networking module, and a gyroscope. The drive circuits of each module are connected to the central processing unit of the robot body.

[0006] The multi-wheel adaptive shock-absorbing chassis includes a chassis, wheels, and at least two sets of rocker arm structures. The chassis is connected to the bottom end of the robot body via shock absorbers.

[0007] The robot body is equipped with a medicine cabinet and a medicine retrieval arm. The motor drives and controllers in the medicine cabinet and the medicine retrieval arm are all connected to the central processing unit of the robot body.

[0008] The medicine cabinet includes multiple liftable independent medicine shelves and magnetic medicine boxes, each with a unique identification code and a magnetic adsorption component.

[0009] The end of the dispensing arm is equipped with a visual recognition unit and a magnetic actuator for accurately adsorbing the medicine box;

[0010] The central processing unit is configured to: construct navigation paths in real time through the environmental perception module and control the movement of the multi-wheel chassis; scan the drug box identification code and patient information through the visual recognition unit to verify drug matching; and trigger an early warning or request path replanning through the communication module when drug misdispensing or path obstacles are detected.

[0011] Furthermore, the multi-wheel adaptive shock absorption chassis is a six-wheel rocker arm type shock absorption chassis, including a chassis, six wheels and two sets of rocker arm structures: the six wheels are fixed to the bottom of the chassis through the frame, each wheel is independently equipped with a steering drive device, and the two sets of rocker arm structures are symmetrically installed on the left and right sides of the chassis and are linked together through a differential.

[0012] Furthermore, the steering drive device is connected to the wheel's rotational support structure via a ball screw pair and connecting rods; the six-wheel rocker arm chassis has a set of rocker arms on each of its left and right sides, and each set of rocker arms is connected to the frame via a connecting rod group and a rotary joint; the maximum tilt angle of the chassis is monitored in real time by a gyroscope, and the wheel speed is dynamically adjusted by the central processing unit to maintain balance. The chassis's motor control device and drive motor are connected to the robot's central processing unit.

[0013] Furthermore, the drug-retrieving arm is a PRPP (prismatic-rotary-prismatic-prismatic) series mechanical arm, which includes a linear motor, a steering servo, a first electric push rod, and a second electric push rod connected in sequence. It is completely decoupled in the three orthogonal degrees of freedom in its workspace. Compared with the traditional RRR (rotary-rotary-rotary) series mechanical arm, the drug-retrieving arm is more accurate in control and positioning, and its code editing process is also simpler.

[0014] The visual recognition unit and the magnetic actuator are located at the end of the second electric push rod. The control circuits of the visual recognition unit and the magnetic actuator are both connected to the central processing unit of the robot body. The magnetic actuator is flexibly connected to the end of the second electric push rod, and the direction of movement is controlled by the central processing unit. The end magnetic actuator is fixedly connected to the end of the second electric push rod and uses a magnetic magnet to attract the medicine box. The attraction force is adaptively adjusted by the central processing unit according to the weight of the medicine box.

[0015] Furthermore, the visual recognition unit and magnetic actuator at the end of the drug-dispensing arm are a camera and an electrically controlled magnetic suction head, respectively, and the control circuits of the camera and the electrically controlled magnetic suction head are both connected to the central processing unit of the robot body.

[0016] Furthermore, the independent medicine shelves of the medicine cabinet are vertically movable via a scissor lift platform; the independent medicine shelves are divided into several independent medicine compartments, and each medicine compartment has a weak magnetic magnet on its back to prevent the medicine box from falling off due to uneven ground when the medicine shelf is raised; the medicine box is magnetically connected to the medicine compartment of the medicine shelf through a weak magnetic adsorption component on its back; the front of the medicine box is equipped with a strong magnetic adsorption component that can be attracted by the magnetic actuator of the medicine retrieval arm and a barcode that can be identified by the visual recognition unit of the medicine retrieval arm.

[0017] Furthermore, the scissor lift platform is jointly controlled by a dual-motor synchronous control system and a servo driver, and both its control circuit and drive circuit are connected to the central processing unit.

[0018] Furthermore, the environmental perception module includes a lidar for generating a three-dimensional environmental map and comparing it with a pre-stored path. When an obstacle is detected, a warning is issued first through the voice module. If there is no response, a detour path is replanned.

[0019] The lidar is located at the upper end of the robot body and includes a laser emitting module, a laser receiving module, a scanning module, and a data processing module. The laser emitting module emits short-pulse lasers towards the target object. After the laser is reflected on the surface of the target object, it is captured by the receiving module. By measuring the time difference between the laser emission and return and the rotation or oscillation of the scanning module, the lidar can generate a detailed three-dimensional point cloud map of the surrounding environment and calculate the distance to the target object by combining the speed of light with the data processing module.

[0020] Furthermore, the central processing unit is equipped with a medical-specific artificial intelligence model to analyze the matching between patient medical record data and drug information, and sends alarm information to a designated terminal through the communication module when an anomaly is detected.

[0021] Furthermore, the upper cover of the magnetic medicine box is connected to the box body through a weak magnetic component, and the inside of the medicine box is provided with an anti-collision buffer layer. The visual recognition unit includes a high-definition camera and a barcode decoder.

[0022] Furthermore, the outside of the medicine dispensing arm is equipped with a glass protective cover.

[0023] Furthermore, the central processing unit is connected to the control circuits of the lidar, the drive device of the six-wheel rocker-arm chassis, the scissor lift platform of the medicine cabinet, and the medicine dispensing arm, respectively. The AI ​​system verifies the matching between the medicine and the patient and dynamically plans the medicine delivery route.

[0024] Furthermore, the display screen is used to show the robot's current remaining battery power, current location, and the ward and bed number of the patient to be visited;

[0025] The positioning module uses a GPS automatic positioning system; the voice module uses a combination of lidar and GPS automatic positioning system to control the prompts issued by the robot.

[0026] The wireless communication module autonomously selects to exchange information with the medical service station based on the GPS automatic positioning system and instructions given by the central processing unit.

[0027] The working principle of the AI-powered medicine delivery robot of this invention for automatic medicine delivery is as follows:

[0028] After the nurse places the medications into the robot's medicine box at the dispensing window and enters the medication name and delivery location into the medicine box barcode, the robot's central processing unit receives the current and target location information signals fed back by the positioning module through the automatic positioning system. It then automatically plans an optimal path, and the robot proceeds to the patient's ward. During its journey, if the lidar detects an obstacle in front of the robot, it will intermittently issue warnings via the voice prompt system. If the obstacle remains within a specified time, the robot uses the lidar to determine if it can detour around it. If it cannot avoid the obstacle, it will use the automatic communication system of the communication module to contact the nearest medical facility. The service platform issues an alert requesting a professional caregiver to handle the situation. This cycle continues until the robot arrives at the patient's ward. The robot scans the patient's medical barcode next to the bed using a camera at the end of its medication arm. It then retrieves the pre-positioned medication box using a scissor-type electric lifting platform, a dual-motor synchronous control system, and a servo drive located below the corresponding medication shelf. The robot scans the barcode on the box and uses an AI system mounted on its central processor to further determine if the medication is correct. If it does not meet medical standards, the robot's automatic communication system sends an alert to the nearest medical service platform requesting a professional caregiver to handle the situation. This process continues until the current medication delivery task is successfully completed and the next delivery task begins.

[0029] Compared with the prior art, the advantages of this utility model are:

[0030] 1. This utility model adopts a unique design of a six-wheel rocker arm shock-absorbing chassis. Through the linkage of the left and right rocker arm structures with the differential, combined with the independent steering drive device and shock absorber, it significantly improves the robot's passability in complex terrain (such as thresholds and slopes), solves the problem of traditional chassis being prone to bumping and jamming, and effectively controls the stability during drug transportation.

[0031] 2. This utility model adopts a PRPP (prismatic-rotary-prismatic-prismatic) serial robotic arm as the medicine-retrieving arm. Through the decoupling design of three-dimensional orthogonal degrees of freedom, it achieves high-precision positioning and simplified control of the medicine box acquisition work. At the same time, the electrically controlled magnetic suction head flexibly connected at the end of the medicine-retrieving arm can adaptively adsorb the medicine box, avoiding the damage to the medicine that may be caused by traditional grippers.

[0032] 3. This utility model uses a weak magnetic medicine rack and magnetic medicine box adsorption design, combined with a scissor lift platform, to ensure the stability of the medicine box during transportation and prevent the medicine box from falling or becoming misaligned. At the same time, a unique drug barcode is set at the front of the medicine box. The barcode is scanned by a camera and compared with the patient's condition barcode. The AI ​​system verifies the drug's matching in real time, filling the gap in the lack of error correction mechanism in the existing technology and further avoiding human error.

[0033] 4. This utility model uses an integrated medicine cabinet, independent medicine rack and independent medicine box, which allows the robot to complete the delivery of medicine to more patients in a single delivery, thereby greatly reducing the robot's operating energy consumption and improving its delivery efficiency.

[0034] 5. This utility model integrates a navigation system (LiDAR + GPS), AI verification, and innovative mechanical structure (chassis + medication dispensing arm). It uses LiDAR to construct a real-time environmental map and a gyroscope to dynamically adjust the chassis attitude. The AI ​​system analyzes patient medical record data using a large language model to proactively warn of medication misdispensing. The magnetic medication box and PRPP-type robotic arm work together to achieve precise medication retrieval and placement. This systematically solves the problems of medication mismatch, transportation bumps, and low medication dispensing accuracy found in existing medication delivery robots.

[0035] 6. This invention connects the central processing unit to each module, analyzes environmental data (obstacles, terrain tilt) in real time, and adjusts wheel speeds in conjunction with the differential, enabling the six-wheeled chassis to autonomously select the optimal path in complex scenarios. Compared to traditional robots with fixed path planning, this significantly improves transportation efficiency and safety.

[0036] 7. This utility model adopts an independent medicine cabinet with replaceable medicine boxes, thereby supporting the classified storage of different medicines. This modular design enables the robot to adapt to the needs of diverse medical scenarios and has greater practicality. Attached Figure Description

[0037] Figure 1 This is a schematic diagram of the overall structure of the robot according to this utility model;

[0038] Figure 2 This is a three-dimensional structural diagram of the six-wheel rocker arm chassis of this utility model;

[0039] Figure 3This is a three-dimensional structural diagram of the robot body of this utility model;

[0040] Figure 4 This is a three-dimensional structural diagram of the medicine dispensing arm of this utility model and the medicine dispensing arm after removing the protective cover.

[0041] Figure 5 This is a three-dimensional structural diagram of the electric push rod at the end of the medicine dispensing arm of this utility model.

[0042] Figure 6 This is a schematic diagram of the operation of the medicine dispensing arm of this utility model when it picks up the medicine box;

[0043] Figure 7 This is an internal cross-sectional view of the medicine box when the medicine dispensing arm of this utility model picks up the medicine box;

[0044] Figure 8 This is a three-dimensional structural diagram of the medicine cabinet of this utility model;

[0045] Figure 9 This is a three-dimensional structural diagram of a single medicine rack according to the present invention;

[0046] Figure 10 A cross-sectional view of a single medicine rack of this utility model with the medicine box loaded;

[0047] Figure 11 This is a three-dimensional structural diagram of a single medicine rack body of the present invention;

[0048] Figure 12 This is a schematic diagram of the back structure of a single medicine rack body of this utility model;

[0049] Figure 13 These are the front view and sectional view of the assembled individual medicine boxes of this utility model;

[0050] Figure 14 This is a three-dimensional structural diagram of the medicine box body of this utility model;

[0051] Figure 15 This is a bottom view of the end cap of the medicine box of this utility model;

[0052] Figure 16 This is a detailed flowchart of the drug dispensing robot of this utility model. Detailed Implementation

[0053] To make the objectives, technical solutions, and advantages of this utility model clearer, the technical solutions of this utility model will be further described below.

[0054] This embodiment proposes an AI-powered medicine delivery robot, such as... Figure 1 As shown, the robot includes a robot body 3, a six-wheeled rocker-arm chassis 1, a medicine cabinet 6, and a medicine-dispensing arm 5.

[0055] like Figure 3 As shown, the robot body 3 has a built-in central processing unit and integrates a display screen 31, an environmental perception module, a voice module, a positioning module, a wireless communication module, a networking module, and a gyroscope. The drive circuits of each module are connected to the central processing unit of the robot body 3.

[0056] The central processing unit is equipped with a medical-specific artificial intelligence model to analyze the matching between patient medical record data and drug information, and sends alarm information to a designated terminal through the communication module when an anomaly is detected. A display screen 31 is located on the outer surface of the robot body 3 to display the robot's current remaining battery power, current location, and the ward and bed number of the patient to be visited. The environmental perception module includes a lidar 4, located at the upper end of the robot body 3, which generates a 3D environment map and compares it with a pre-stored path. When an obstacle is detected, a warning is issued first through the voice module; if there is no response, a detour path is replanned. The positioning module uses a GPS automatic positioning system to confirm the robot's relative position in space in order to dynamically plan the robot's next running route. The voice module controls the robot in conjunction with the lidar 4 and the GPS automatic positioning system. The human-generated prompts are used to issue pre-set, designated prompts to the outside world under specific circumstances. The wireless communication module autonomously selects to exchange information with the medical service station based on the GPS automatic positioning system and instructions from the central processor. Working together with the networking module, it sends an alarm to the nearest medical service platform via wireless signal when the robot is in a special environment or abnormal state, requesting professional caregivers to take further action. The gyroscope is used to detect the current posture of the robot body. When the robot's tilt angle is too large, it can prevent the transported drugs from being severely jolted or damaged by reducing the speed of the drive wheels or replanning the route.

[0057] In this embodiment, the lidar 4 is threadedly fixed to the upper end of the robot body 3. It includes a laser emitting module, a laser receiving module, a scanning module, and a data processing module. The laser emitting module emits short-pulse lasers towards the target object. After the laser is reflected on the surface of the target object, it is captured by the receiving module. By measuring the time difference between the laser emission and return and the rotation or swing of the scanning module, the lidar can generate a detailed three-dimensional point cloud map of the surrounding environment and calculate the distance to the target object by combining the speed of light with the data processing module.

[0058] like Figure 2As shown, the six-wheel rocker arm chassis 1 includes a chassis, six wheels, and two sets of rocker arm structures. The chassis is connected to the bottom end of the robot body 3 via four shock absorbers 2 through threads. The six wheels are fixed to the bottom end of the chassis via a frame, and each wheel is independently equipped with a steering drive device. The two sets of rocker arm structures are symmetrically installed on the left and right sides of the chassis and are linked together through a differential. The steering drive device is connected to the wheel's rotational support structure via a ball screw pair and connecting rod. Each side of the six-wheel rocker arm chassis has a set of rocker arms, and each set of rocker arm structures is connected to the frame via a connecting rod group and a rotary joint. The maximum tilt angle of the chassis is monitored in real time by a gyroscope, and the wheel speed is dynamically adjusted by the central processing unit to maintain balance. The chassis's motor control device and drive motor are connected to the central processing unit of the robot body.

[0059] like Figure 4 As shown, the medicine-dispensing arm 5 is threadedly connected to the support plate of the robot body 3. It is a PRPP (prismatic-revolute-prismatic-prismatic) series mechanical arm, comprising a linear motor 51, a steering servo 52, a first four-section electric actuator 53, and a second four-section electric actuator 54, all threaded together in sequence. This allows the second four-section electric actuator 54 to move horizontally and vertically along the end face of the medicine cabinet 6, and to rotate around the first four-section electric actuator 53 or extend along the positive and negative directions of the robot body 3. The mechanical arm is completely decoupled in its three orthogonal degrees of freedom within its workspace. Compared to the traditional RRR (revolute-revolute-revolute) series mechanical arm, the medicine-dispensing arm 5 in this embodiment offers greater control and positioning accuracy, and its code editing process is also simpler. The medicine-dispensing arm 5 is also equipped with a glass protective cover 55, which can be made of carbon fiber composite material or ceramic material with good heat dissipation properties.

[0060] In addition, such as Figure 5 As shown, the ends of the second four-section electric push rod 54 are respectively equipped with a camera 542 and an electrically controlled magnetic suction head 541. The control circuits of the camera 542 and the electrically controlled magnetic suction head 541 are connected to the central processing unit of the robot body 3 for precise grasping of medicine boxes. The electrically controlled magnetic suction head 541 is flexibly connected to the end of the second four-section electric push rod 54 and its movement direction is controlled by the central processing unit. The electrically controlled magnetic suction head 541 is fixedly connected to the end of the second electric push rod 54 and uses a magnetic magnet to attract the medicine box. The attraction force is adaptively adjusted by the central processing unit according to the weight of the medicine box.

[0061] like Figure 8As shown, the medicine cabinet 6 is fixed to the support plate of the robot body 3 by a threaded connection. The medicine cabinet 6 includes a medicine cabinet body 62 and six liftable independent medicine shelves 63. The medicine cabinet body 62 is attached with a current medicine cabinet barcode 61. Each independent medicine shelf 63 contains a magnetic medicine box. Each medicine box 631 is equipped with a unique identification code 6311, a magnetic adsorption component 6312 and a weak magnetic adsorption component 6313 on the back.

[0062] like Figure 9-11 As shown, the independent medicine shelf 63 of the medicine cabinet 6 includes a shelf body 632 and a scissor lift platform 633. The scissor lift platform 633 is fixedly installed at the bottom of the shelf body 632, enabling smooth vertical movement. The scissor lift platform 633 is jointly controlled by a dual-motor synchronous control system and a servo driver, with both its control and drive circuits connected to a central processing unit. Simultaneously, the independent medicine shelf 63 is divided into 36 independent compartments for storing medicine boxes, such as... Figure 12 As shown, each medicine compartment has a weakly magnetic ferrite magnet 6321 on its back to prevent the medicine box from falling off due to uneven ground when the medicine rack 63 rises; each medicine box 631 has a weakly magnetic attracting material on its back that can be attracted to the medicine rack. In this embodiment, for example... Figure 13-15 As shown, the magnetic material on the back of each medicine box 631 is made of a weakly magnetic ferrite magnet 6313, which magnetically attracts the medicine box into the compartment; as Figure 6-7 As shown, the front interior of a single medicine box 631 is equipped with a strong magnetic neodymium iron boron magnet 6312 that can be attracted by the magnetically controlled magnetic suction head 541 of the medicine dispensing arm, and the front exterior is equipped with a current medicine box barcode 6311 that can be recognized by the medicine dispensing arm camera 542. This magnetic medicine box 631 includes a box body and an upper cover 6314. The upper cover 6314 is magnetically connected to the ferrite magnet 6315 of the box body via a weakly magnetic ferrite magnet 63141, and the inside of the medicine box is equipped with an anti-collision buffer layer.

[0063] In this embodiment, the motor drives and controllers in the medicine cabinet 6 and the medicine dispensing arm 5 are both connected to the central processing unit of the robot body 3.

[0064] In this embodiment, the central processing unit is connected to the control circuits of the lidar 4, the drive device of the six-wheel rocker-arm chassis 1, the scissor lift platform 633 of the medicine cabinet 6, and the medicine dispensing arm 5, respectively. It verifies the compatibility of the medicine with the patient through an AI system and dynamically plans the delivery route. Its configuration includes: constructing a navigation path in real time using the lidar 4 and controlling the movement of the six-wheel rocker-arm chassis 1; scanning the identification code 631 of the medicine box (i.e., the current medicine box barcode 6311) and the patient information barcode using the camera 542 to verify the compatibility of the medicine; and triggering an early warning or requesting route replanning through the communication module when a mis-dispensing of medicine or a path obstacle is detected.

[0065] like Figure 16 As shown, the basic process of automatic drug delivery by the AI ​​drug delivery robot in this embodiment of the utility model is as follows:

[0066] Each module's drive and controller are connected to the robot's central processing unit (CPU). When the nurse places medications into the robot's medicine box 631 at the dispensing window and enters the medication name and delivery location into the medicine box barcode 6311, the CPU of the robot body 3 automatically plans an optimal path based on the current and target location information fed back by the positioning module (automatic positioning system). The robot then proceeds to the patient's ward. During this process, if the lidar 4 detects an obstacle in front of the robot, it will intermittently issue a warning via the voice module (voice prompt system). If the obstacle remains within a specified time, the robot uses the lidar 4 to determine if it can detour around it. If it cannot avoid the obstacle, it will communicate with the communication module and network module (automatic positioning system). The communication system sends an alert to the nearest medical service platform, requesting a professional caregiver to handle the situation. This process continues until the robot arrives at the patient's ward. Upon arrival, the robot scans the patient's medical condition barcode next to the bed using the camera 542 at the front end of its medication arm 5. It then retrieves the pre-positioned medicine box 631 from the corresponding medicine shelf 63 via the shear-type electric lifting platform 633, dual-motor synchronous control system, and servo drive, and scans the barcode 6311 on it. The AI ​​system mounted on the central processor further determines whether the medication is correct. If it does not meet medical standards, the robot's main body 3 sends an alert to the nearest medical service platform through its automatic communication system, requesting a professional caregiver to handle the situation. This process continues until the current medication delivery task is successfully completed and the next delivery task begins.

[0067] The above are merely preferred embodiments of this utility model and do not constitute any limitation on this utility model. Any equivalent substitutions or modifications made by those skilled in the art to the technical solutions and contents disclosed in this utility model without departing from the scope of the technical solutions of this utility model shall still fall within the protection scope of this utility model.

Claims

1. An AI-powered medicine delivery robot, characterized in that, Includes the robot body and a multi-wheel adaptive shock-absorbing chassis: The robot body has a built-in central processing unit and integrates a display screen, an environmental perception module, a voice module, a positioning module, a wireless communication module, a networking module, and a gyroscope. The multi-wheel adaptive shock-absorbing chassis includes a chassis, wheels, and at least two sets of rocker arm structures. The chassis is connected to the bottom end of the robot body via shock absorbers. The robot body is equipped with a medicine cabinet and a medicine dispensing arm. The medicine cabinet includes multiple liftable independent medicine shelves and magnetic medicine boxes, each with a unique identification code and a magnetic adsorption component. The end of the medicine dispensing arm is equipped with a visual recognition unit and a magnetic actuator for accurately grasping the medicine box; The central processing unit is configured to: construct navigation paths in real time through the environmental perception module and control the movement of the multi-wheel chassis; scan the drug box identification code and patient information through the visual recognition unit to verify drug matching; and trigger an early warning or request path replanning through the communication module when drug misdispensing or path obstacles are detected.

2. The AI ​​medicine delivery robot according to claim 1, characterized in that, The multi-wheel adaptive shock absorption chassis is a six-wheel rocker arm type shock absorption chassis, including a chassis, six wheels and two sets of rocker arm structures: the six wheels are fixed to the bottom of the chassis by the frame, each wheel is independently equipped with a steering drive device, and the two sets of rocker arm structures are symmetrically installed on the left and right sides of the chassis and are linked by a differential.

3. The AI ​​medicine delivery robot according to claim 2, characterized in that, The steering drive device is connected to the wheel's slewing support structure via a ball screw pair and connecting rods; each rocker arm structure is connected to the frame via a connecting rod group and a rotary joint; the maximum tilt angle of the chassis is monitored in real time by a gyroscope, and the wheel speed is dynamically adjusted by a central processing unit to maintain balance.

4. The AI ​​medicine delivery robot according to claim 1, characterized in that, The dispensing arm is a PRPP type series structure, which includes a linear motor, a steering servo, a first electric push rod, and a second electric push rod connected in sequence, and is completely decoupled in three orthogonal degrees of freedom in its workspace; the vision recognition unit and the magnetic actuator are located at the end of the second electric push rod, and the magnetic actuator is flexibly connected to the end of the second electric push rod, and the movement direction is controlled by the central processing unit; the end magnetic actuator attracts the medicine box through a magnetic magnet, and the attraction force is adaptively adjusted by the central processing unit according to the weight of the medicine box.

5. The AI ​​medicine delivery robot according to claim 1, characterized in that, The independent medicine rack of the medicine cabinet can be moved vertically by means of a scissor lift platform; the independent medicine rack is divided into several independent medicine compartments, and each medicine compartment is equipped with a weak magnetic magnet on the back; the medicine box is magnetically connected to the medicine compartment of the medicine rack through a weak magnetic adsorption component on the back; the front end of the medicine box is equipped with a strong magnetic adsorption component that can be attracted by the magnetic actuator of the medicine dispensing arm and a barcode that can be identified by the visual recognition unit of the medicine dispensing arm.

6. The AI ​​medicine delivery robot according to claim 5, characterized in that, The scissor lift platform is jointly controlled by a dual-motor synchronous control system and a servo driver, with both the control circuit and the drive circuit connected to the central processing unit.

7. The AI ​​medicine delivery robot according to claim 1, characterized in that, The environmental perception module includes a lidar, which is used to generate a three-dimensional environmental map and compare it with a pre-stored path. When an obstacle is detected, a warning is issued first through the voice module. If there is no response, a detour path is replanned.

8. The AI ​​medicine delivery robot according to claim 6, characterized in that, The central processing unit is equipped with a medical-specific artificial intelligence model, which is used to analyze the matching between patient medical record data and drug information, and to send alarm information to designated terminals through the communication module when an anomaly is detected.

9. The AI ​​medicine delivery robot according to claim 1, characterized in that, The upper cover of the magnetic medicine box is connected to the box body through a weak magnetic component, and the inside of the medicine box is equipped with an anti-collision buffer layer. The visual recognition unit includes a high-definition camera and a barcode decoder.

10. The AI ​​medicine delivery robot according to claim 4, characterized in that, The medicine dispensing arm is equipped with a glass protective cover.

Images